Okay, everybody. Once again, welcome to the 2024 MicroCap Rodeo, brought to you by Lucosky Brookman and The Money Channel. I'm John Heffernan, one of the moderators today. I do some work over at Madison Square Garden as one of their announcers, so it's a pleasure to be here. If you've heard my spiel in the last presentation, apologies. For those of you who are new, welcome. After this fifth session today, we'll be taking a one-hour break for lunch. But I want to introduce our next presenter, and he, originally Pennsylvania guy, now in Princeton, New Jersey, the CEO of the Mistras Group. Let's hear it from Mr. Ed Prajzner. Eddie! What did I say?
CEO. That's my wish.
I'm just reading what they give me, man.
All right. Good, good job. Thank you, John. Appreciate it.
This is premature.
There you go.
Yeah.
Thank you. All right, thank you. I know I stand between the group here in the room and lunch, so I'll keep my comments kind of brief here and keep us right on schedule. Appreciate your time and interest, and appreciate the sponsors in being able to bring us together today to do a little deep dive on the Mistras Group. So, first and foremost, before I go too far here, I will just give you the quick little legal forewarning here that I will be making a few forward-looking statements using EBITDA, non-GAAP terms, things like that. So consider yourself forewarned, and all those such information is appropriately reconciled in our filings, which you can look through at your leisure. This whole presentation is on our website as well.
I flipped the order of a couple of pages just for the easier presentation, but you'll find all this content on our webpage. So, real high level, I'm going to assume this group has, you know, fairly rudimentary knowledge of who Mistras is. I'm going to kind of go from that level, but I will dive into some of the newer, exciting things we're really evolving into as well. So if you have a pretty advanced knowledge, I'll touch on a few things as well for your benefit and information here. So who we are at a real high glance here. We are an asset testing company. We are a nondestructive asset testing company. That's what we do. We do it with various techniques, processes, proprietary. It's people-oriented, it's equipment-oriented. Data is a big part of that story.
We use sensors as well as our own proprietary equipment, but there's lots of ways we do that. Why we're doing it is basically for safety is one key thing. We are sort of an ESG enabler for my customer. I want them to keep uptime, minimize their downtime, while they're using very mission-critical assets in industrial environments, and I'll walk you through who they are. Again, we are based in Princeton, New Jersey. We went, did our IPO back in 2009. And, we've got a hundred and, you know, twenty-some locations at this moment and, just around 5,000 employees. Many of them are highly trained technicians in the field that are gathering the data and doing the trend analysis for my customers. Right now, we have a very motivated team.
We're being led at the moment by Manny Stamatakis, who's Chairman of the Board and Interim CEO. I am CFO, and you've got the other names there, John Smith, Jerry D'Alterio, Hani Hammad are the heads of operations now. Today, Sotirios Vahaviolos is our founder, and a lot of the technologies that I talk about, he was the author of that, the creator of that, and much of that is still in the DNA of how we operate today, and I'll walk you through some of that in a minute. First, though, let me take one quick step back. What is this NDT thing that I'm talking about? It's nondestructive testing. I'll also maybe say nondestructive evaluation, with an E on the end.
What that basically is, is I want to give my customer some state of their assets, the condition, the integrity of the assets. But I can't impair the integrity while I do that. It has to be minimally or non-intrusive here, along the way. If I rip open the side of a piece of equipment to show them there's no corrosion inside, great, but now I have a bigger problem. I've just impaired the asset itself by creating a problem. So we don't want to punch into the equipment. I have to do it from afar. So what you're seeing on the page here is basically how I do that. It might be just my eyes. Maybe I'm a skilled technician. I can tell right away that that's good or bad, but it goes much deeper than that very quickly.
Maybe it's an ultrasonic test, maybe it's an X-ray or some other technique where I want to get a much clearer image of what I can't see in front of my two eyes here, to give you, the customer, the comfort that the asset is in a good condition there. So a lot of this is, and it goes to much more advanced techniques. Again, starts with a human, starts with a piece of equipment. It goes more advanced into, well, maybe a sensor needs to collect more data over time, and then it gets into data. Maybe I start to trend the data, create an algorithm to help you figure out what's happening. So it's assets at the beginning of their life cycle, while they're being commissioned, all the way to the end of life being decommissioned.
I'll test them along the way, and they're being tested against a standard, essentially. But it's all about the asset integrity, keeping it up in its intended condition, minimizing repairs and maintenance, and hopefully pushing off expensive capital repairs as long as you can, is another purpose and goal here. But safety is really job one, two, and three, why I'm out there testing these assets for the customer. So that's what non-destructive testing is. Where do I do it? Is the next thing I just want to give you sort of a quick highlight here. Lots of end markets, but it's heavy industrial users of assets. Think of power plants and refineries, windmills, bridges, it's places like that is where we're working.
Those types of environments where assets are under physical stress, load, pressure, temperature, you name it, corrosion, natural elements, that's the asset base that I'm looking at. And I do it, the service offering that I give is in a couple of different ways. So on the shop service, sorry, in the services side, I'm going to test the equipment in the field. I'm going to a power plant, to a refinery, to a wind farm, to test the asset right then and there. Now, that carries its own set of challenges. Maybe it's at height, getting access to the site, being trained to be on those assets, carries a whole, you know, sort of, precursory set of things to do to get out there. Much of my revenue is there now.
That's kind of where we got started along the way. The other extreme would be my shop laboratory. That's the part coming to me to be tested. Maybe it's a chunk of titanium going on to a jet engine motor. I'll be testing that part along the way. So if the part's small enough, comes to me, that's generally an OEM part, or I may go visit it in the field if it's already installed. In the middle there, and where I'll spend much of the time today, is on the data side, where it ties both together, the field and the shop, both can use data. But that's a smarter proposition in a way, because now I don't have to necessarily come look at the asset.
I can interpret, infer, what's happening, the state it's in, and I'll walk you through what that means and examples of that, as we go through a few more, few more slides here. But the tech is important here in the field, critically important. The trained technicians, we believe we have, you know, one of the, if not the largest trained, nondestructive technician workforce in the US It does start and end with this person. They visit the customer frequently, high connectivity. They have a lot of tribal knowledge. They've been out inspecting, you know, similar common assets for quite some time. A lot of knowledge there. Safety is a reason they're there, front and center, and they are very embedded with the customer.
My techs essentially become an outsourced or co-sourced asset integrity manager with my customer, deeply embedded with them, helping them manage the risk of their assets. So again, technicians are highly trained, working at height with multiple certifications in ultrasonic techniques, radiological techniques, et cetera. And they're not a lone op. Technicians aren't just doing difficult, back-breaking work, no doubt, but we try to automate what they do. So over time, here's a few examples of that. So I'm in the field. That top picture there is a robotic crawling device taking an X-ray, believe it or not. That's a little device that we built in-house, and it's proprietary and patented to us. So that's going to speed along. That's above ground pipe in that case.
So as opposed to a tech in the past, lifting up a piece of equipment, taking an X-ray every 18ft or so, now that little device spins along with a collar, takes a 360 image, and moves along the length of pipe, creating a more holistic view of what's happening on that asset, versus haphazardly testing as a human would have. So that's a huge advantage there. Moves much quicker than a human could have, and it's giving the customer more, you know, robust information about the state of their asset. That crawler can also work over insulated pipe. That's in that case. It's not insulated pipe, but it can see through what's going on underneath. Again, the whole trick there is seeing what you can't see with your naked eyes and giving the customer a great view of that.
So that's an example of an above-ground piece of equipment. The other picture there is a buried pipe. So in that case, it's a PIG, as I affectionately call them, pipeline inspection gauges. That PIG would run underground through the pipe. Similarly, not taking you know x-rays in this case, but looking for you know the strength of the metal, the integrity. Is there pinholes that shouldn't be there? Is the weld okay? The ovality, is the pipe being squashed from pressure, and it's not quite round? Things like that is what it's checking for. So highly technical, and that PIG can run you know a few hundred feet or a few thousand or miles at a time. So that's all regulatory things.
In that case, the PHMSA folks, who are part of the US Department of Transportation, mandate that that testing happen. So that's, you know, generally in the midstream space, which is a less volatile sector of oil and gas that we like. But again, this creates efficiency and productivity gains for the technician; these devices will be used in the field. Again, in the shop laboratory testing environment, the part comes to me. Here's a simple example. I won't get into much detail here, but to the far left there is a foundry where a casting, an extrusion, you know, some raw piece of metal, in this case, we'll use titanium, comes to me. Before it's machined, before it's manufactured, before I apply overhead to it, I wanna see.
My customer wants to see, is there a void in there? Was it cast properly? Did the foundry do their thing? If there's a void inside of that metal, send it right back and melt it down now and start over. Don't apply all this overhead to find a non-conforming part down the line. So we do that step up first, the basics, and we've done that for quite some time. But more importantly now, because my commercial aerospace and private space and defense customers are clamoring for it, I'm doing more and more steps in my own shop laboratories now. Surface treating it, putting a small notch in it.
A little CNC machining goes a long way to help accelerate that part down the line to get it ready, in this case, to be assembled as a fan blade on a jet engine motor for commercial aerospace aircraft. In the past, all these steps were done by multiple vendors. Each vendor did a step, had to come back for testing, so it extended out the supply chain. By me taking on all these steps, I can speed up that final delivery to final assembly for the customer and improve the quality along the way. If you listen to our earnings calls of the last, you know, two, three, four of them, we talk a lot about this. We're putting more CapEx here.
Some of my customers are graciously lending me capital and equipment to expand my capabilities here, because you know, commercial aerospace, private space, and defense are all really good sectors, growing nicely now, and I can really leverage my shop labs here into second and third shifts and really speed through the testing here. The best part, too, is unlike the field, where I'm more time and materials, this is more per part, per pound, per linear foot kind of testing. So the more efficiency I gain there, the more upside for me and my economics. But lots of good partnership there with the customer on the shop side. Data. So I'm gonna kinda spend the rest of the time on the data piece here, which really ties everything together.
We've been in data for quite some time, from the very beginning. We've been selling software for quite some time, but we never really brought it to the forefront as we are now. And we've talked a lot about this. If you go back our last, you know, two or three or four earnings calls, we're talking more about this. It's gaining more momentum, and it's evolving, you know, quickly here. Because we're really realizing this is a sort of an underserved piece of our market and, you know, we've sold data to some customers and never sold them service, and we sold other customers service, never sold them data. Either one can pull the other one forward. Data can pull through service, and service can pull through data.
And why this is really important, because in the field or in the shop, you need that, and we do capture the data. We can accelerate that through. We have all the other capabilities here as well, but it's a very, very important piece of the puzzle. One. I'm gonna go one slide forward and come back. A big hallmark of this is right smack in the center of our data solution is called PCMS software, Plant Condition Management Software. We acquired this from BP many years ago, and what this software does, and this is used in just over half of the US refineries today use this software. And what it's basically allowing them to do is more risk-based look at what they're looking at.
This is a simple heat map of consequence and probability of damage on their assets. So they're looking at an asset class. Refineries have every asset known to mankind, piping and valves and vessels, et cetera. So each one of them can be looked at in this manner. So obviously, the red ones would be high probability of an incident and high consequence of damage. I care most about those assets. I don't want them to be having me stay up all night worrying about them, so I really want a robust testing plan for them. The green assets, low consequence, low probability, I don't probably need to test there as often. I might be over-testing those assets. The ones in the yellow and the amber, I need to dial in and approach there.
But first and foremost, I need to think about what's the most, what, what's gonna hurt me? Let's go there first and really better, you know, think about the risk here in a more meaningful way at the asset level. That's what PCMS is doing. This is tried and true and has been out there for quite some time. This used to be more of a proprietary bricks-and-mortar application. We've more recently gone to a SaaS model. Much easier to bring this up for the customers. We've put this under SOC 2 recently. SOC 2 is an audit comfort provision on my software, so my customers can very easily hook this into their enterprise systems and embed it in their own operations. So this is a core of my data going forward. I'm gonna go back there. Sorry. But it's just one piece of the puzzle.
On the nine o'clock position is what we just looked at, PCMS. It's only one piece of the puzzle. So after I look at my PCMS kind of risk management, you know, matrix, I then go up to that twelve o'clock position and think about, okay, what is the testing plan? What more information do I need? What insights do I want? How do I want to help them manage these assets? But that's just a plan. I now have to go out, and now I'm showing you the chart before, the pig and the crawler. I have to go back out and check this once in a while, because maybe somebody changed the crude type or the catalyst or the material type changed.
I wanna go validate that, capture some test measurements, bring them back in, and maybe I need to monitor longer term, is what that, kind of that four o'clock position is telling me. And then I bring that data back in, whether a technician took a reading and tapped it on a tablet or a sensor did it or a device did it, bring it back into PCMS, warehouse that data, and then I can trend and compare and analyze what's going on. Why is that metal thickness, you know, accelerating, you know, deteriorating more aggressively than I thought? Maybe there's something I can do about that. Maybe they changed the metal type. Maybe they changed the catalyst. You can learn from that.
Again, they don't wanna have an unplanned outage, they don't wanna extend out a planned outage, and they certainly don't wanna have premature capital repairs that they could have avoided. So this whole sort of bundled solution, this one-source solution, it is a real value for the customer. There's much greater ROI in that. Plus, it connects me to the customer. I've got my equipment involved here, my technician, my software, but the real, you know, benefit of this is it gets you into this whole RBI testing, risk-based inspection, away from simply time-based testing. Because maybe I kept sending my technicians out to haphazardly test, you know, assets in the green all the time. That's not a real effective use of their time and money. This testing is expensive.
So let's dial in and go after the most prone assets for failure with the most intelligence here, really spreading the you know use this data I'm collecting for the customer's benefit to help them de-risk their environment and maximize safety, maximize uptime, and minimize the things that could be going wrong in their sites. That's what really ties it all together, but you know it really starts with data as a forefront of that. A metal thickness reading would be a central piece of data here in this particular example, and relevant to many, many of the environments that we're testing. So again, PCMS is critical to it. This is a huge opportunity for us. There's a real you know accelerated growth path we can get to here.
Data is about 10-ish% of my revenue today. We believe it can grow, you know, nicely over time. It's also rather focused right now on my core markets. Down below, a little hard to see, but the bolded areas, oil and gas, and then other process industries, that's where my data is today. That's where PCMS is primarily. All the other unhighlighted areas there, aerospace and defense, infrastructure, industrials, petrochem, on and on, I have not really, you know, brought the data to them.
I do lots of NDT services there, and that's a very kind of easy upsell, cross-sell to them to bring that data to them, and that's where we're going to expand that first and foremost, but lots and lots of assets being managed right now under PCMS, and lots more we can expand to just tangentially adjacent to the end markets we serve right now. One example of this, to really kind of illustrate this, and again, I know lunch is coming, so I will cut us off here shortly. But real example of how this operates, I can't quite give you the customer name here. Unfortunately, that's confidential, but on the left side there, real example. This is a customer who spends $50 million a year over every two-year cycle doing turnarounds at their facility.
So they're breaking down assets, turning off assets, shutting down systems. You're doing very extensive testing and capital repairs and maintenance, et cetera, and lots of NDT testing, NDT testing to see what's going on. That's their nut right now they have to deal with. Well, we brought PCMS to the table to risk-based how they're testing, why they're testing, when they're testing, and we were able to save 10% of that effort. So $50 million, they saved $5 million of that. They'll save that every two years. Significant amount of effort goes away there, for a one-time fee of $500,000, in this case. 10-to-1 payback in the first year for using PCMS, they're gonna get that savings going forward forevermore, and they can learn and expand over time.
I won't go through the example on the right side there. A little more complicated, but same concept there. Save them a lot of downtime and complexity. So huge upside there, huge advantage for the customer. Now, you might say, "Well, well, why did you do that? Because you were involved in some of that turnaround testing. You were also doing some of the services in that $50 million, you know, number there. You just cut some of your own work down, didn't you?" Maybe I did. Maybe I cannibalized a little bit of my effort, but I'm more than happy to do that because I'll pick up a higher margin on the data I'm selling to them now as a subscription, as a service going forward. More importantly, I'm gonna create all this connectivity with them.
Now they're gonna use PCMS, so that little cutaway chart, you know, that we talked about, now I'm gonna have this dashboard with the customer all the time. I'm gonna know what they're doing. I'm now gonna know how to recalibrate. It's a barrier to entry. My competitors don't have this bundle that I have. You know, there's more ROI I can demonstrate to them. I can expand this to their other sites. So it just creates a much stronger relationship, a higher ROI, a win-win with the customer and a higher partnership. And my technicians maybe don't have to do a haphazard test now. Maybe they'll be out there calibrating a sensor. Once in a while, you have to still collect more data to make sure the dataset is full and algorithms know every fact and instance that's possible.
So there you know, maybe the technician will do a light mechanical fix now while they're there, instead of giving the customer a headache that they're compelled to act on now. Instead of finding a problem, the tech could now still do a test, but say, "Hey, it's not in a bad state. It's even better condition than the algorithm or PCMS would have otherwise implied." That's a good news thing versus giving them sort of unknown issues popping up while they're there doing a test. So it just creates a more holistic relationship, more value add, with higher levels of kind of predictability and expectations when we're doing the work, versus, "Hey, I spent more time doing the testing," and, "Hey, you have more problems here than you thought.
Have a nice day," and I leave with a punch list for the customer. That's not helpful, versus I stayed there with a sensor out front, giving me rudimentary data, tech showing up at the right interval, not at any old - not at a scheduled time, but at the right exact time to stay in front of things that's preventing that little crack in the metal from propagating into failure. I can see it sooner, interact and have the customer intervene before it gets to a fail-safe position. So that's more value for me there, that I avoided the downtime, either unplanned downtime or I minimized the scheduled downtime. Best of all, the CapEx, you know, can last a little longer here. They're not replacing things prematurely. Again, safety is job one, two, and three here.
We lead off all of our internal safety meetings on that front, and that's a big reason why we are there. So I know we're running short, and I know I'm about to get yanked off the stage here, and lunch is coming, but you know, this presentation is on our website. You can contact me on the email and phone number there anytime you want. Happy to explain more about Mistras and who we are and where we're going, and again, appreciate your time today in following the story. Sure, I could certainly field one or two from the audience here, or if anyone wants to maybe explore something more or two? Yes, sir.
[inaudible] traditionally have a lot of volatility and visibility with the revenue, and I think oil and gas, you know, a lot of sensitivity. Has that changed, and how have you kind of addressed that so that visibility is improved, et cetera?
Yeah, great question about our visibility. In the past, if you go back, you know, five-plus years ago, there was a lot of volatility, a lot of seasonality, wherein refineries in particular, they would have done two big turnarounds, a spring turnaround, a fall turnaround. We had two peaky spiked quarters where the revenue was ginormous, and it settled back down the other two quarters. Now, over time, with this data model I'm talking about, we're more embedded with the customer. So we're doing smaller turnarounds all year long, staying embedded with the customer, so it smooths the year out now. So if you looked at my four quarters, they look very ordinary. There's a little tiny seasonality now to maybe two and three, but it's not a wild spike now. So yes, that gives me, we call it run-and-maintain business.
So I have these long-term evergreen contracts governed by master service agreements, where I like to stay on site. My same group of technicians stays out on site, many times visiting the same customer all year long. So it kind of really helps kind of smooth out the load over the course of a year, because it's really hard to staff and for that one peaky period of time when everybody needs every tech and every piece of equipment I can find, very difficult to do. So over time, we kind of managed this with data, with software, with scheduling the workload differently, many cycles going down throughout the whole course of a year versus one big bang when everybody went offline. That was a very bad model for everybody.
So over time, we've worked consciously to smooth that out, you know, and have more predictability of what's happening on our end markets and be able to staff up and serve that, because that was very hard to catch the peak, because you had to carry this big bench in the down cycles. Very hard to deal with that when you wanted to catch the peak. So you had to give up on the peak sometimes because you couldn't carry the bench during the low periods. So very challenging to pull that off. So the industry, I think, collectively, has learned that that was not a great economic model, so that has smoothed over time.
So I think the net result of that is we have a little more predictability, more line of sight, a little easier to schedule up and staff for those kind of cycles. But again, data goes a long way as well, because now I'm not running blind from the last test I was out there. I have some other data along the way that can patch together what's happening to those assets. That goes a long way as well. So you're not doing-
Got it.
A little, definitely, a SaaS model. Our software definitely is. So then you're not really flying blind and trying to figure out a big gap from point A to point B. You know, kind of what's happening in the middle there. Goes a long way as well to minimize sort of haphazard testing when people might show up and have to reconnect the story. They get more connected all the way along there. But a lot less volatility than we've had historically at the current time. Yep. Yes, sir.
Can you talk a little bit about the competitive dynamics there for the services you offer? Can you characterize sort of your kind of market rank, market share, major competitors?
Sure, sure. I mean, I would generally categorize it as a fairly decentralized market, so there's lots and lots of competition, fairly decentralized. If you added up the couple of large competitors in this space, that would be Mistras, Team Industrial, Acuren. Even adding up the big three, as I would call us, you still would only get a 10-15% market share at most, probably. There's lots and lots of other smaller competitors. What I would add, though, is very few competitors cover the gamut as we would cover, with the technician, the product, the sensor, the software. We're very unique in that regard. We believe our offering is very broad, but it is very competitive, and in many cases, my customer tries to do the work and self-perform for themselves.
So a lot of our work is an overflow, where the customer. And it's very hard to keep, you know, ultrasonic-trained technicians on a staff or radiological-trained techs. They don't have the, you know, workload to balance that out. So they, you know, it ebbs and flows a little bit, where sometimes the customers do rudimentary testing. They would augment with our staff. So the customer, ironically, not that we compete with them, but that's a big place of where some of our share comes from. But, no, we, it's a highly competitive market, you know, and, you know, we openly bid on work. A lot of it's long-term nature. You do keep customers for quite some time.
In many cases, you know, you collaborate with competition to catch the peaks at a customer site, where sometimes we will be partnering with one of our competitors, just joining forces to catch the peak of a workload at a customer site. So there's a little more collaboration, cooperation sometimes in our industry as well. But no, it's very highly fragmented, and no one has a significant share of any sort here. And there's plenty of work to do, and the market is growing nicely here for the needs for our services on the data side, on the shop side, even in the field side, you know, as people want to extend out the asset lives of equipment in the field, you know, there is a growing demand for our services.
Okay. Could you say anything about market share for the past five years? The same market share as far as in,
Yeah, a good question. A little hard to say. I'd say over the last five years, I would say probably at least holding serve and holding share. Over the last year, I think we're gaining some share. Our growth rates, I believe, have been a little greater than our competition over the last year or so. We've consciously leaned in there with more commercial efforts to really grow on that side. We've got a little more strategic on the pricing side as well. So I think our revenue is growing a little greater than the industry right now, but the market is rather significant, and you know, it is growing across the board.
Taking work from customers is a place we can kind of grow that hidden market that we don't compete with today, but have the customer trust us to give us the work. This is an area where we see a lot of growth coming from. As well as other vendors get involved, we take work away from adjacent vendors that might work up and downstream of our testing, and we take on an additive step right before or after the nondestructive test. That's another area where we can grow, and we are in the commercial aerospace, private space, and defense world. Taking on adjacent, you know, pools and markets that we didn't traditionally compete in is another area where we can grow.
That shop business, where we have this aerospace business, has been growing close to 20% for the last couple of quarters, if not year now. So that, that's an area where there's a lot of exceptional growth going, and, you know, we'll keep taking share there.
[inaudible] your shareholder base? I read somewhere on the aggregation there that it's about 40% or insider. I don't know.
You do not quite 40, a little lower, probably 35. If you have a fairly large insider holding between directors and officers, yes, you're up in probably the 35% range there, so you have a fair amount of insider holdings there today. Yep.
Final question for you. Do you ever, with your data solutions in your earnings call sheets, have you ever tossed out the words artificial intelligence? Because that could sometimes seem to [inaudible]
Yeah. Good. Yeah. No, no, good question. Yeah, it's tempting, but probably not quite, 'cause there's so much involved in these assets and humans involved and the risk. I don't think you'd ever take the human exactly out of this equation, 'cause there's too much at risk, and safety is involved and regulations. An algorithm, yes, to boost up the intelligence of the human, but I don't think you can take the technician out of this equation entirely. So we want to move in that direction, but I think I would stop quite a little bit short of that in this regard. But, yeah, I understand where you're coming from. So, thank you. I appreciate all the good questions. Great, great questions, so I appreciate your time and being able to present a little bit about the Mistras story to you today. Thank you.
Thank you.
All right, thank you. Appreciate it. Thank you.
Folks, once again, that's Senior EVP and CFO, Mr. Ed Prajzner of Mistras. Folks, Ed, thank you. Great job. Thanks for coming out, everybody. We're gonna take a break from 12:00 P.M. to 1:00 P.M. for lunch, and here in track one, 1:00 PM, we'll kick off our afternoon agenda, starting with GoHealth and their VP of IR and speaker, John Shave and Vijay Kotte. We'll see you then, everybody. Enjoy lunch, and thank you once again, Ed Prajzner.